**Voca: Local-First Text-to-Speech for GrapheneOS**

GM & PV

#Introductions

Voca is a text-to-speech application, built by cypherpunks for cypherpunks.

Voca is free and open-source, designed exclusively for GrapheneOS. It operates fully offline with zero Google dependencies.

You can queue and render text instantly, no tracking, no ads, no compromises. Just text read aloud for you.

Development is funded exclusively by donations: the more you zap, the more we ship.

Voca is just getting started, so if you want to see this app in reality, let us know we're on the right track with a fat zap to pay the devs.

Reply to this note

Please Login to reply.

Discussion

Hi voca, and welcome to #nostr! πŸ€“

βœ… Great, you've already set up your #lightning address! I just ⚑️zapped you some #welcomesats 😊

I can also help people check their #nostr address is working (if they have one). Nostr addresses are not mandatory, but some people like using them to add a personal touch to their profile. Once you've set it up, just post a note including #checkmyaddy and I'll check it again for you. If you like, I can give you one for free at "yourname@minestr.app". Just get in touch! 😊

Made with πŸ’š by nostr:npub1mlnersd9cup2n3p5xrytccgfa73p4resxe5a2ljq9rez2ancr8usmnu7x8

How do I integrate Amethyst with it?

There shouldn't be any integration needed, the app reads any text shared with it.

The flow would be: copy text > share > voca read now

I want it to keep watching my feed and read every new post it sees as a desktop companion. How do we do that?

Voca is scoped to be android only, at least at first.

It could potentially take an rss feed or npub and wait for new events to be added to the queue.

Amethyst is Android only as well. So, I could just send every post to you as soon as we get one.

Voca uses sherpaTTS under the hood, maybe you can just use that directly? I remember a play button in older amethyst versions that did TTS.

This is very interesting.

We're looking for kotlin devs, so please DM me if you're interested to help out!

No problem!

nostr:nprofile1qqsyvrp9u6p0mfur9dfdru3d853tx9mdjuhkphxuxgfwmryja7zsvhqpzamhxue69uhhv6t5daezumn0wd68yvfwvdhk6tcpz9mhxue69uhkummnw3ezuamfdejj7qgwwaehxw309ahx7uewd3hkctcscpyug please any ideas already for integration?

No idea. But I am open to interesting ways to integrate. We use TTS to say people's names on their profile. But that's it right now.

I see. One idea I had was integrating with DMs, or integrating with replies to one's posts(reading them out loud).

Yeah, that or reading the feed would be very cool. We just need to figure out a good interface for it.

Maybe even reading notifications...

Yeah!

I'd like to have this workflow:

- I find an interesting article (that I don't want to read on mobile phone)

- copy - > share to some nostr note

- voca listens, converts it to audio, dumps it to a podcast RSS feed

- I go and listen to it in Fountain at my own pace sometimes later

*something like this would make sense in the Highlighter as well.

YES

That's the third major feature on our roadmap: on any text/pdf/website, click share, convert to mp3, in the background voca runs the TTS engine, generates a mp3 and places it in a folder where your podcast app catches it and adds it to the queue.

We are still many dev hours away from that, so please zap if you want to see this happen sooner!

nostr:npub1m2mvvpjugwdehtaskrcl7ksvdqnnhnjur9v6g9v266nss504q7mqvlr8p9 does this with his blog and it's a neat idea.

https://fountain.fm/show/3MjdPUN26O8yQAtaXBmL

loaditfor.me is awesome!

And since the TTS engine is on a powerful server, the audio quality is better than what Sherpa can do on the phone.

πŸ”₯πŸ”₯πŸ”₯

This must have been one of those tabs I opened to "read more carefully later" and it slipped away!

Just tested it out with njump link of a note, it works like butter. Just requires a lot of manual "spreading" - clicking. But what a way to nail a simple MVP, nostr:nprofile1qqsd4dkxqewy8xum47ctpu0ltgxxsfemeewpjkdyzk9ddfcg286s0dsppemhxue69uhkummn9ekx7mp0qywhwumn8ghj7mn0wd68ytnzd96xxmmfdejhytnnda3kjctv9uq36amnwvaz7tmwdaehgu3wvduhq6r9wfc82mnt9e6x7erp0yhs4deh46!

It would be a killer feature if I can do it w/ a single click directly from the client or a browser. Definitely willing to pay with zaps or any other seamless method.

* I didn't find the podcast feed in Fountain search, but AntennaPod got it quickly.

** I just saw Ryan Hoover resurfacing same product need on the bird app, and some inreresting discussion about relqted features https://xcancel.com/rrhoover/status/1904841208960373041

One click would be nice, although not sure how to implement it.

In fountain it should be searchable, but you can always add private rss feed.

I'll be open sourcing the code for both rss2podcast and loaditfor.me soon. I was experimenting and rewriting it with my intern a lot (ok, it was mostly vibe coded), but now it has become more stable.

Btw I'm "consulting" a bit on nostr:nprofile1qqs0tj5e4fj8ljvy0tdkrdhdyqxek9vsjwjmjunvfe8zhxdlnlfj8rcppemhxue69uhkummn9ekx7mp0qy2hwumn8ghj7mn0wd68ytn00p68ytnyv4mz7fudpxl .

Big fan of the project 😁

Yeah this link is from the Fountain search. https://fountain.fm/show/3MjdPUN26O8yQAtaXBmL

Anything added to podcastindex will show up in any app that uses it for search.

This would be great for CalyxOS too.

It should work on any android, but grapheneOS is the prime target.

Speak button is disabled, on graphene

Do you have a text to speech engine installed?

We recommend SherpaTTS, get it on nostr:nprofile1qqs83nn04fezvsu89p8xg7axjwye2u67errat3dx2um725fs7qnrqlgzqtdq0

Didn't have it the first time. After installing sherpatts from zapstore, it's still not speaking

Open Sherpa and download an engine, we recommend en_GB-southern_english_female-medium in the piper models.

It worked. Thanks man πŸ‘

That's awesome!

It's kinda clunky to have to setup the TTS engine in a second app, but better do one thing right.

Might add some tutorial on how to setup Sherpa though.

nostr:nprofile1qqs0tj5e4fj8ljvy0tdkrdhdyqxek9vsjwjmjunvfe8zhxdlnlfj8rcppemhxue69uhkummn9ekx7mp0qy2hwumn8ghj7un9d3shjtnyv9kh2uewd9hj7qg4waehxw309ahx7um5wghx77r5wghxgetk9undltsg needs funding and devs:

Users already lined up πŸ™‹πŸ™‹πŸ™‹πŸ™‹

Zaps go all to Voca

nostr:nevent1qqs2e6aqn7hwfmgpqmv83h6mt2j80lqw6d024vfup0jzd3fdtulnt0spr4mhxue69uhkummnw3ezucnfw33k76twv4ezuum0vd5kzmp0qgs0tj5e4fj8ljvy0tdkrdhdyqxek9vsjwjmjunvfe8zhxdlnlfj8rcrqsqqqqqpdfypt7

Your post leads with "Voca is a text-to-speech application..." but then at the end of your post it sounds like it's just an idea, not an app (which would explain why I couldn't find it in the stock F-Droid repos). What is the current status?

Very early stage work in progress.

Currently gauging if there is demand, which seems to be the case.

Can I choose a specific celebrity's voice?

Voca works with the local text to speech engine, there are a handful of different voices available in SherpaTTS.

πŸ‘€

It's made for freaks like you!

t *Y*

Cumulative zaps: 210 000 sats

Cumulative expenses: 1 200 000 sats

So far voca made 1 mil sat loss.

Let's see if we can get profitable by zaps alone.

#v4v

nostr:nevent1qqs2e6aqn7hwfmgpqmv83h6mt2j80lqw6d024vfup0jzd3fdtulnt0spzamhxue69uhhyetvv9ujumn0wd68ytnzv9hxgtczyr6u4xd2v3lunpr6mdsmdmfqpkd3tyyn5kuhymzwfc4en0ul6v3c7qcyqqqqqqgs97nwz

Check this out nostr:npub1smkgazuplxhcr2ggekpsmeglqk08nfdteg8fh98xpae7n97enmcs89ctp3